agentleFS
RomeCoreRomeCore/LLMDesktopAssistant/.agents/skills/idea-to-ticket/SKILL.md

idea-to-ticket

Skill1 starsChanged 33 days agoMITView on GitHub

Record an idea and the agent's proposal for it as one small ticket under docs/ideas/<scope>/. Use when the user shares a feature idea, a design preference or a "would be nice" and a proposal comes back to it, when the user asks to write the idea down / file it / remember it / turn it into a ticket, or when a decision from a discussion deserves to outlive the chat. This is an idea-and-proposal store, not a plan, a spec or a wayfinder map.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. qepo17/skills★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository11 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the idea turns out to be genuinely big, say so and let the user pick the heavier vehicle.

    line 27
  2. Distinctive:

    Create the folder on first use.

    line 37
  3. Distinctive:

    Write the user's side only from what the user actually said.

    line 80
  4. Distinctive:

    English, neutral spec prose: No chat quotes, no colloquialisms, no swearing, no first person, no "as we discussed".

    line 84
  5. Distinctive:

    Prefer the project's domain vocabulary over invented names.

    line 87

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Idea to ticket (dASS)

A ticket here is a two-voice record: what the user wanted, and what the agent proposed back. It exists so an idea and the thinking around it outlive the chat window — not to schedule work.

When it is active

  • The user shares an idea, and a proposal comes back to it.
  • The user says "write this down", "file it", "make a ticket", "remember this", "put it in docs/ideas".
  • A decision, a correction or a refused alternative appeared in the discussion and is worth keeping.
  • The user invokes /skill:idea-to-ticket.

What this is not

NotBecause
A planNo stages, no ticket graph, no per-stage checklist, no issues/ subfolder. One file, one idea.
A specIt records a discussion, not a locked design with numbered requirements.
A wayfinder mapNo research/prototype/grilling children, no frontier, no claim-and-resolve cycle.
A bug reportSomething behaving unlike its description goes to dass-report.

If the idea turns out to be genuinely big, say so and let the user pick the heavier vehicle. Do not quietly grow this format into one.

Where it goes

docs/ideas/<scope>/<idea-slug>.md — one idea per file, kebab-cased slug, no numbers.

docs/ideas/ is the idea and proposal store and sits deliberately outside the issue tracker: nothing is dispatched from here, and no frontier scans it.

  • <scope> groups ideas by topic (web-ui, lua-api, ...), taken from the skill argument or from the conversation. Create the folder on first use; a scope is a topic, never a plan.
  • The header borrows the tracker's vocabulary (docs/issues/README.md is its authority) so an idea can be promoted into docs/issues/ without being reworded.

The format

# <Short title of the outcome, not of the topic>

Status: ready-for-agent
Type: task
Blocked by:

## Requested

- <the user's idea, point by point, in requirement language>
- <each constraint or preference they stated>
- <the open question they named, if they named one>

## Proposed design

- **<Claim.>** <Why, and what it replaces or corrects.>
- <what is accepted from the request as it stands>
- <what is added, corrected or refused, and on what grounds>

## Not in scope

- <the tempting adjacent thing, and why it is not being taken>

## Comments

Optional sections, when they carry weight: ## Prerequisite (something that must exist first), ## Alternatives considered (when a refused option needs its reasons kept).

Type: task for work-shaped ideas. Status: uses the tracker's vocabulary: open while something blocks it or the design is unsettled, ready-for-agent when it is settled and nothing does, resolved once done — plus the triage strings. Blocked by: names another idea when one genuinely gates the other, and stays empty when nothing does.

Two sections, two voices

Keeping them apart is the whole point of the file.

  • ## Requested is the user's voice, paraphrased into requirements. Their asks, their constraints, the shape they proposed, the question they left open. Requirements language, never a transcript.
  • ## Proposed design is the agent's voice. What is accepted as it stands, what is corrected, what is added, what is refused and on what grounds. An agent that only agrees is not doing this job: the corrections are the value, and so are the refusals — they stop the same argument from being had twice.

Write the user's side only from what the user actually said. Do not answer their questions on their behalf, and do not upgrade a half-formed idea into a decision they never made.

House style

  • English, neutral spec prose. No chat quotes, no colloquialisms, no swearing, no first person, no "as we discussed".
  • No file paths, no code. They go stale; name types and concepts instead. Exception: a shape that encodes a decision more precisely than prose — a marker contract, a state machine, a type sketch — trimmed to the decision-rich part.
  • Small. Bullets, one claim per line. A ticket that needs a table of contents is the wrong vehicle.
  • Prefer the project's domain vocabulary over invented names.

What bites

  • Blending the voices — the requested and the proposed sections answering each other line for line, so the reader cannot tell the idea from the opinion about it.
  • A plan in disguise — stages, cross-ticket dependencies, checklists, a nested issues/ folder.
  • Quoting the conversation — the user's phrasing, in their language, verbatim. Paraphrase into requirements.
  • Inventing a Status: string that is not in the tracker's vocabulary.
  • Filing the proposal while dropping what it refused — the refusals are half the record.

Not here

  • Deciding whether an idea deserves work at all → grilling.
  • Slicing an approved design into implementable tickets → to-tickets.
  • A bug about dASS itself → dass-report.
  • Turning a finished session into reusable addons → equip.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.